🧠 Open to Learn Together

Karasik

SubMaroon

·

AI & ML interests

tg: @kar_asik Current plans: MLP train 26B-A4B; OV_proj + MLP train 31B; merge new base model for training

Recent Activity

updated a model 37 minutes ago

SubMaroon/Rumeni-26B-A4B

published a model about 7 hours ago

SubMaroon/Rumeni-26B-A4B

updated a model about 7 hours ago

SubMaroon-exp/Rumeni-26B-A4B-heretic-GGUF

View all activity

Organizations

upvoted a paper about 1 month ago

Functionality-Oriented LLM Merging on the Fisher--Rao Manifold

Paper • 2603.04972 • Published Mar 5 • 4

upvoted a paper over 1 year ago

SynthDetoxM: Modern LLMs are Few-Shot Parallel Detoxification Data Annotators

Paper • 2502.06394 • Published Feb 10, 2025 • 89